Synopsis At about 1610 on 23June2002, the amphibious vehicle LadyDuck took on water while on the Ottawa River during a combined land and water-borne sightseeing tour of the National Capital Region. The vehicle sank rapidly by the bow in eight metres of water when near the Hull Marina. Of the 12people on board, 6passengers, the driver, and the tour guide escaped from the vehicle and were recovered by private craft on the scene at the time of the sinking. Four passengers, trapped within the sinking vehicle, drowned. There was no environmental damage. Ce rapport est galement disponible en franais. Minister of Public Works and Government Services 2004 Cat.No.TU4-14/2004E-HTML ISBN0-662-36856-8 1.0 Factual Information 1.1 Particulars of the Vehicle 1.1.1 Description of the Vehicle The LadyDuck was an amphibious vehicle arranged to carry up to 12passengers on combined road and water-borne tours in the National Capital Region (NCR)3 and on the Ottawa River. The vehicle was developed and built by the owner and entered commercial service at the start of the tourist season in June2001. The vehicle was based on the conversion of a FordF-350 truck chassis (seePhoto1). The original gasoline engine was used for on-road operation. A gasoline-powered Mercruiser inboard/outboard (I/O) motor at the rear was used for water-borne propulsion. Figure1 shows the basic layout of the vehicle. Photo1. LadyDuckafter recovery at the Hull Marina Figure1. Construction profile outline The bottom of the chassis was enclosed with welded and bolted steel plating and the sides were extended upward to enclose a buoyant structure. The original truck wheels and suspension were fitted outside the watertight hull. The effective breadth of the chassis was increased by the addition of extensions (sponsons) on each side. These sponsons were partially filled with rigid foam plastic to enhance transverse stability and buoyancy and to ensure appropriate forward and after trim when water-borne. A passenger boarding ramp, located at the rear left corner of the vehicle, hinged up to a steel sill and a flexible gasket to ensure the watertight boundary of the passenger deck. The ramp was lifted to the closed position by an electrically driven winch and secured in place with hasps on each side of the ramp. Eleven single-passenger seats were arranged along the sides of the open passenger deck, with five on the port side and six on the starboard side. There were two seats in the vehicle cab, the left one for the driver and the right one for the tour guide. In the event of a full payload, the tour guide's seat acted as the 12thpassenger seat; the tour guide would assume a standing or crouching position in the middle of the cabin aisle. A fabric awning provided overhead protection from inclement weather. Roll-down transparent weather screens were arranged on each side of the passenger area for additional protection. Approved lifejackets were stowed in lockers under each of the passenger seats and, at the time of the occurrence, three children's lifejackets were located at the after end of the vehicle. Twelve additional adult-approved personal flotation devices (PFDs), located adjacent to the seats in the passenger area, were readily available. A steel visor plate was fitted at the front of the forward engine compartment and was intended to prevent the entry of water at the front of the vehicle when water-borne and underway. Combustion and cooling air supply to the road engine were provided by two triangular ventilation openings in the top of the hood and also by a pair of 76mm-diameter (3-inch) cowl-shaped ventilators fitted at the forward end, above the top of the visor. The vehicle was fitted with five drainage points, each comprised of a hand-operated, ball check valve (seacock) with a swing-check, non-return valve immediately inboard, both of which were mounted on a common pipe spigot and closed with a screwed steel plug. Four of the drainage points were 19mm (inch) in diameter; the fifth was 12mm (inch) in diameter. Once on shore, the drainage points could be used to release accumulated water from the bilges (seeFigure1). The vehicle was equipped with six electrically driven submersible bilge pumps. One pump of 1250 US gal/h (78.8 L/min ) maximum rated capacity and another of 630USgal/h (39.7L/min ) maximum rated capacity were located in port and starboard bilge drain wells that extended below the hull bottom near the mid-length of the vehicle. These two pumps were referred to as .the main bilge pumps. Two other pumps, each of 1250USgal/h (78.8L/min) maximum rated capacity, were located in the hull, slightly forward of the bilge wells, and were referred to as emergency bilge pumps.4 A submersible bilge pump of 1100USgal/h (69.4L/min ) maximum rated capacity, incorporating an automatic float switch, was located at the forward end of the forward engine compartment. Another submersible bilge pump of 630USgal/h (39.7L/min) maximum rated capacity, also with an automatic float switch, was located in the Mercruiser I/O engine compartment at the after end of the vehicle. A manual bilge pump was installed on the port side of the vehicle abaft the driver's seat, with a flexible overboard discharge hose. All electrically driven bilge pumps, except that in the I/O engine compartment, were operable from the steering position. No automatic bilge high level alarms were fitted. The vehicle's communication aids included a very high frequency (VHF) radio, a public address system, and a two-way radio, all adjacent to the driving position. 1.2 History of the Voyage 1.2.1 On-land Operation of Vehicle On Sunday, 23June2002, the driver went to the company's yard and, in accordance with standing instructions, conducted an inspection and other preparatory procedures to prepare the LadyDuckfor tour operation. He then drove the vehicle to the Hull Marina where, at about 0820, he conducted the water-borne operational test. On completion of the test at 0835, the vehicle was driven to the company kiosk on Sparks Street in Ottawa. At about 1030, having completed an all on-land tour, the driver was informed that another company vehicle (LadyDiveIII) had broken down and that the LadyDuckwould carry out the next amphibious tour. 1.2.2 First Amphibious Tour of the Day The water-borne part of the tour commenced at approximately 1100. During routine operation of the main bilge pumps, no water was seen to be discharging from the bilge outlets near midships. The driver switched on the emergency bilge pumps and observed water intermittently discharging on each side of the vehicle near midships. The driver notified the company's mechanic of the situation by two-way radio. It was suggested that the bilge pumps be switched off for a few minutes, then restarted. This action was unsuccessful and the water-borne tour continued with the emergency bilge pumps in continuous operation. Upon returning to the kiosk, the driver called the mechanic again for further instructions. He was directed to inspect the fuse box, where it was found that both main bilge pump fuses were burned out. These fuses were then replaced with two new 20-amperefuses. 1.2.3 Second Amphibious Tour of the Day Before departing on the second amphibious tour of the day, the driver was instructed by the company manager to include more on-land and less water-borne time due to increased traffic on the river, such that approximately 15minutes would be added to the on-land portion of the tour and the water-borne portion would be reduced proportionately. The water-borne part of the tour was to take the vehicle slightly upstream of the Alexandra Bridge, then downstream of the Macdonald-Cartier Bridge close enough to the Rideau Falls for the passengers to see them, before returning to the Hull Marina ramp. The LadyDuck started the amphibious tour at about 1500, with the driver, 10passengers, and a tour guide on board. At the beginning of the tour, the guide briefed the passengers, in French and English, on safety procedures related to the on-land part of the tour. Before the vehicle entered the water at the Hull Marina ramp at approximately 1540, the tour guide provided a safety briefing to the passengers for the water-borne part of the tour. When the vehicle entered the water, the main bilge pumps were switched on to clear the hull of any shipped water. Because no water was seen to be discharging from the outlets near midships, the emergency bilge pumps were also activated to discharge the accumulation of floodwater. Water was then seen discharging intermittently from outlets on both sides of the vehicle near midships. The bilge pump, located in the forward engine compartment, was also switched on manually and a red light adjacent to the steering position indicated that the pump was supplied with electrical power. The switch was then placed in the automatic operating position and the pump was considered to be active. Once the vehicle was afloat, the front engine was kept running in neutral, the I/O motor was started, and the water-borne part of the tour commenced. The vehicle was driven to the Ottawa side of the river at approximately eight kilometres per hour (km/h) and then at slower speeds to various points of interest so that passengers could take photographs. The weather was fine and clear with little wind. The river was relatively calm, with waves caused by wakes from boats and other watercraft in the tour area. On occasion, the vehicle encountered waves that washed over the hood and up to the windshield. Some spray also came in through the opened windows in way of the driver and tour guide seats. Toward the end of the tour, at about 1608, while returning to the Hull Marina at approximately 8km/h, the driver noted that the front end of the vehicle was floating lower than normal and that water was being continuously discharged from both sides of the vehicle near midships. The driver then ordered the four foremost passengers and the tour guide to move to the back of the vehicle to try to decrease the forward trim. The forward trim continued to increase and, realizing that the safety of the passengers was at risk, the driver instructed the tour guide to tell the passengers to don PFDs. At this time, he diverted the vehicle toward the nearest point on the Quebec shore. The driver then broadcast a MAYDAY on emergency VHF channel16, identifying the LadyDuck,giving its position and the number of passengers on board. At about 1610, the situation deteriorated rapidly as more floodwater accumulated in the forward end of the vehicle. The driver then called on the passengers to abandon the sinking vehicle. The driver left the steering position, made his way aft and, with seven other persons, managed to get free of the sinking vehicle. The remaining four passengers became trapped under the fabric awning and sank with the vehicle in 8m of water. Figure2. Chart of the occurrence area The driver, the tour guide, and four of the passengers who escaped as the vehicle was sinking were quickly recovered by the pontoon boat LePirr'eau, which happened to be near the LadyDuck as it sank. Two more passengers were recovered by the pleasure craft MarinadeHull,which had responded promptly after witnessing the sinking. The bodies of the four passengers who sank with the vehicle were recovered later and were subsequently examined by the coroner. The coroner determined the cause of death to be drowning associated with hypothermia. Following recovery from the water, the survivors were taken to the Hull Marina office for first-aid medical attention; those with no apparent injuries were subsequently released. The driver and two passengers, requiring further medical attention, were transported to the Ottawa Hospital and released later that day. 1.3 Injuries to Persons 1.4 Search and Rescue Response The driver's radio distress call on emergency VHF channel16 was received by some of the pleasure craft in the area. However, the driver of the LadyDuck had no time to acknowledge any of their responses before abandoning the vehicle. Several bystanders with cellular telephones called the 911emergency telephone service and informed emergency services in Hull5 and Ottawa of the occurrence. Ottawa and Hull fire and police emergency response services were alerted between 1612 and 1615, and were on site by1620. At 1635, Marine Communications and Traffic Services (MCTS) Prescott were advised of the occurrence by the pleasure craft CatchtheSun. MCTS Prescott relayed this information to MCTS Qubec. The Hull police and fire departments launched their rescue craft shortly after 1620 and the Hull police chief of operations assumed the duty of on-scene commander. As Hull fire department divers were not equipped to descend 8m to the sunken vehicle, the services of the Ottawa Underwater Recovery Team Unit were requested. At approximately 1800, the Underwater Recovery Team Unit recovered the bodies of four passengers from the sunken vehicle. All the victims were found at the rear of the LadyDuck,floating against the underside of the awning. Two adult victims were recovered wearing adult PFDs. Before the sinking, the two younger victims were wearing adult PFDs; however, on recovery, one was found wearing a partially donned adult PFD and the other was recovered without a PFD. 1.5 Vehicle Recovery 1.5.1 Recovery Operations The sunken vehicle was located in 8 m of water, approximately 90m from the Quebec shore and 120m east of the Hull Marina ramp in approximate position 4526.1'Nlatitude, 7542.3'Wlongitude. The LadyDuck was facing in a westerly direction, in an upright condition with its front wheels and propeller drive unit turned to the right. Divers subsequently straightened the front wheels to facilitate ramp haul out. Following extensive underwater videotape recording of the vehicle, including its fittings and the positions of all bilge pump energizing switches, the LadyDuckwas recovered on 27June2002 and transported to the TSB Engineering Laboratory in Ottawa for further inspection. 1.5.2 Vehicle Condition on Recovery The propeller drive unit of the I/O motor was found to be in the raised position. The boarding ramp at the after end of the open passenger space was found to be in the raised and fully secured position on the watertight gasket surrounding the access opening. The five bilge drainage valves were found in the fully closed position with their steel screwed plugs in place. The fabric awning over the open passenger space was in place and secured to its metal frame. One roll-down transparent weather screen on the starboard side near midships and the foremost weather screen on the port side were securely zipped in the lowered position. The foremost weather screen on the starboard side and two other weather screens on the port side were undone and hanging loose. The remaining roll-down transparent weather screens were secured in the raised position. 1.5.3 Vehicle Examination after Recovery After recovery, the partially flooded vehicle leaked freely from several hull penetrations, and was more rapidly cleared of floodwater by removing the screwed plugs from the bilge drains and opening all five seacocks. On boarding the vehicle, the following items were noted: flare gun, flares, and portable air horn, first-aid kit and flashlight, boating safety kit, emergency steering gear handle for I/O motor, three fire extinguishers, one Danforth anchor and anchor rope, mechanic tool kit, mooring lines, two paddles, spare motor oil, booster cables, logbook, pre-departure vehicle checklists, and spare fuses (30 and 15 amperes). flare gun, flares, and portable air horn, first-aid kit and flashlight, boating safety kit, emergency steering gear handle for I/O motor, three fire extinguishers, one Danforth anchor and anchor rope, mechanic tool kit, mooring lines, two paddles, spare motor oil, booster cables, logbook, pre-departure vehicle checklists, and spare fuses (30 and 15 amperes). 1.6 Passenger Seating Passenger seats were arranged on the port and starboard sides of the vehicle, with a centre aisle giving access fore and aft. Metal lockers, forming the base of each seat, were fastened to the 12mm-thick (-inch) plywood floor by 3mm-diameter (1/8-inch) screws. The original Ford truck driver and passenger seats were retained in the steering cabin. The number of screws securing the tour passenger seat bases varied from a minimum of three up to a maximum of six. The plywood flooring and metal seatings in way of some of the screwed connections were in a deteriorated condition. Transport Canada's (TC) Motor Vehicle Safety Regulations, Standard207, requires seat anchorages to withstand a force 20 times the weight of the empty seat (20g6), applied in a longitudinal direction through the centre of gravity of the seat. TSB tests and calculations indicate that a seat from the LadyDuckanchored with four fastenings would require a force of 9.7g to separate from the flooring (seeAppendixG - List of Supporting Reports). 1.7 Bilge Pumping Arrangements 1.7.1 Bilge Pumps Installation Photo2. After end of the starboard-side main bilge pump discharge pipe The starboard-side electrically driven submersible main bilge pump (AttwoodV1250), located and secured in a bilge well near the mid-length of the vehicle, was fitted with a fine wire gauze strainer. The discharge piping was comprised of various lengths of flexible 25mm-diameter (1-inch) concertina-type plastic hose, 20mm-diameter and 16mm-diameter (-inchand5/8-inch) rubber hoses, and 12mm-diameter (-inch) metal ferrule joining pieces. The discharge piping led aft along the starboard side of the passenger area and passed through the transom some 240mm (9inches) above the load waterline, creating an operating pressure head of some 900mm (35inches) at the pump. The after end of the flexible discharge pipe protruded outside the transom, extended some 305mm (12inches) below the load waterline, and was out of sight of the driver (seePhoto2). An overboard discharge fitting, adjacent to the pump and some 100mm (4inches) below the gunwale, was blanked off, having become redundant when the original bilge pump discharge hose was re-directed aft and through the transom. The on/off switch that energized the pump was located in an overhead panel adjacent to the driving position. The port- and starboard-side electrically driven submersible emergency bilge pumps (both Attwood V1250), each located approximately 610mm (24inches) forward of the bilge wells, were not fitted with wire mesh strainers nor were they secured to the bottom hull structure. Each pump was fitted with a 25mm-diameter (1-inch) flexible discharge hose, led to an 11mm-diameter (7/16-inch) discharge fitting located 150mm (6inches) below the gunwale, creating an operating pressure head of some 1070mm (42inches) at the pump. The locations of the overboard fittings on each side near the mid-length of the vehicle were such that they could be seen from the driver's position in the left- and right-hand side rearview mirrors. The on/off energizing switches were located in an overhead panel adjacent to the driving position. The electrically driven submersible bilge pump (AttwoodV625), located in the after engine compartment, was equipped with a float switch. The pump was energized locally by a switch arranged for manual and automatic setting. A 20mm-diameter (-inch) flexible hose, some 610mm (24inches) long, was attached to the pump discharge but was not led overboard, so that any bilge or floodwater issuing from the pump would be retained inside the vehicle. An electrically driven submersible bilge pump (RuleMate1100), incorporating a float switch, was located at the forward end of the forward engine compartment. The pump was actuated from a panel located below the steering column by a switch arranged with automatic, off and manual settings. The manual switch was spring-loaded and must be held in that position by the driver for the pump to continue operating. A 25mm-diameter (1-inch) flexible concertina-type discharge hose led up over the top and forward of the visor plate and the hood flexible gasket, some 460mm (18inches) above the load waterline. The hose extended some 200mm (8inches) below the load waterline at the front of the vehicle and was out of sight of the driver (seePhoto3). When the hood was closed, the discharge hose was compressed where it passed over the top of the forward sealing gasket, partially restricting the flow of any water issuing from the pump. Photo 3. Discharge hose from the forward engine compartment A 38mm-diameter (1-inch) manual piston pump (BecksonThirsty Mate136PF6), located on the port side abaft the driver's seat, was supplied with a flexible discharge hose capable of being led overboard over the top of the gunwale. The pump suction extended to within 12mm (inch) of the bottom shell plating. The pump barrel below the plywood flooring was locally compressed; however, the pump operated freely, and the discharge rate was dependent on the arm strength and stamina of the operator. A pump suction hose, with a 20mm-diameter (-inch) end connection flange, elbow and attached strainer, was located in way of the foremost passenger seat base on the port side of the vehicle. The suction hose penetrated the plywood flooring and extended down into the hull to within 12mm (inch) of the bottom shell plating. This suction became redundant when a small gasoline-driven bilge pump, to which it had been connected, was removed from the vehicle some time before the occurrence. Figure3 shows the location of the bilge pumps. Table1 gives a summary description of the bilge pumps and outlines their status. See AppendixA for results of bilge pump examination and testing. 1.7.2 Bilge Pump Condition The quoted maximum rated discharge capacity is generally that which is attainable at zero pressure head. The actual operational capacity depends on several parameters related to the installation, which include the following: operating discharge pressure head, length and diameter of discharge hoses, diameter and location of overboard discharge, unrestricted pump suction, and maker's installation instructions. Of the six electrically driven bilge pumps on board the vehicle at the time of the occurrence, two were found to be effectively operable. Actual discharge rates measured during the tests of the two emergency bilge pumps (AttwoodV1250) that were operational were 335USgal/h (21.2L/min) each. This discharge rate was approximately 36percent of that specified for this pump model when operating at a comparable pressure head of 3.3feet (1m) with an electrical supply of 13.6volts(V). The reduction of the designed discharge rate was attributable, in part, to the actual voltage available during the tests (approximately12.5V), mud and debris obstruction of the wire gauze strainer, and some internal wear in the pump. However, the primary cause of the reduction was due to the greater back pressure at the pumps, caused by the marked and sudden restriction of the discharge area to 19percent of the original 25mm-diameter (1-inch) pump discharge piping where it joined the 11mm-diameter (7/16-inch) overboard discharge fitting. Further inspections showed that the electrical circuits of the AttwoodV625 andV1250 main bilge pumps incorporated 15-ampere and 20-amperefuses, instead of 2-and4-amperefuses as indicated on each of the respective pump housings and specified in the maker's Installation Instructions (FormNo.69439,Rev.B). Furthermore, these instructions also included the following information: ! WARNING : Always use the fuse amperage rating specified for your pump model. Failure to do so could result in serious personal injury or fire hazard. Attwood pumps are designed to exhaust STANDING WATER ONLY. They are not intended to prevent rapid accumulation of on-board water due to rough weather, hull damage, and/or other unsafe navigational conditions. WIRING INSTRUCTIONS : - - - - - NOTE : Failure to make waterproof connections and fuse pump properly will void the product warranty. (sic) The bilge pump wiring and other electrical service circuits included connections that were hand-twisted and wrapped with electrical tape. The wiring connections did not incorporate soldered and heat-shrunk watertight tube fittings, as in generally accepted marine and automotive practice. The main bilge pumps were secured to manufacturer-supplied horizontal mounting brackets that were located in the bilge drain wells. The emergency bilge pumps and those fitted at the forward and after ends of the vehicle were unattached to any structural member. These pumps were free to move or fall over in the event of shocks due to uneven road surfaces or the sudden surge of floodwater inside the hull. 1.7.3 Bilge Pumping Requirements Bilge pumping requirements for small passenger vessels are set out in the Marine Machinery Regulationsand the Small Vessel Regulations (SVR). With respect to the LadyDuck, the Marine Machinery Regulationsrequire at least two effective bilge pumps and the SVR require one bailer and one manual bilge pump. However, the regulatory requirements were designed for small commercially operated vessels of conventional construction and layout and are tacitly based on the premise that vessels have an acceptable level of watertight integrity and possess adequate freeboard for the intended service. The LadyDuck, which was not of conventional construction, was equipped with six power-driven bilge pumps and one manual bilge pump. At the time of the sinking, two of the power-driven pumps were effectively operable. No bilge high level alarm was fitted in the LadyDuckand none was required by regulation. 1.8 Vehicle Operational Trials 1.8.1 Trials Preparation and Sequence A series of trials was conducted by the TSB to determine the trim, freeboard, wave making, and other operational characteristics of the LadyDuckthroughout a range of speeds, wave heights, and flooding conditions to provide a basis on which the operational condition of the vehicle at the time of the sinking could be assessed. The trials were conducted with three TSB personnel on board the vehicle, which was loaded with sand bags to simulate a typical fully loaded operating condition. The steel visor plate at the front of the road engine compartment was fixed in the fully raised position, and the hood sides, engine ventilation openings and forward cowl ventilators were temporarily sealed weathertight. The flexible foam gasket in way of the hood and visor top at the forward end of the forward engine compartment was retained in its original in-service condition. The Mercruiser I/O motor at the after end of the vehicle was made operational for the in-the-water trials, but the road service engine at the forward end was not restored to working order. The original outfit of damaged or inoperative electrically driven bilge pumps was replaced with new pumps of similar models and capacities, but the original bilge piping system and discharge fittings were retained. The starboard-side main bilge pump discharge pipe protruding from the transom was raised clear of the water to prevent any syphonic flooding throughout the speed trials. Plywood flooring in the passenger space was removed to provide a clear view of the internal hull structure, drive shaft bearing, bilge pumps, and piping systems. Freeboard measuring frames were installed at the forward and after ends of the vehicle, with a crossbar on the forward frame located level with the top of the fixed internal visor plate. The frames were installed to provide clear external indications of the fore and aft trim and effective freeboards during the speed, bow wave, and flooding trials. Photographic, video, and documentary records of the trials and tests were maintained by TSB personnel located on shore and on board the vehicle. Vehicle inspections and all in-the-water trials were attended and observed by representatives of the Ministry of Transportation of Ontario (MTO), TC, the Quebec coroner's office and the owner. These representatives were also supplied with copies of the results of the following trials: Hull integrity, trim and freeboard tests Speed, bow wave and forward freeboard trials Bilge piping syphonic action trials Ramp launching trials Floodwater and trim at various forward freeboards 1.8.2 Hull Integrity, Trim and Freeboard Tests While the vehicle was being launched at the start of the trials, a significant quantity of water entered the hull by way of the main drive shaft bearing. The condition of the bearing was as it was when the vehicle was recovered, having been replaced three days before the occurrence. In order to reduce the inflow of water and allow the trials to continue, the vehicle was hauled clear of the water and the bearing was fully charged with grease, pumped in through a piped lubrication fitting located at the outboard end of the bearing housing, and accessible from outside the watertight boundary of the hull. Due to on-road operation, the drive shaft bearing (FafnirmodelRCJ-1-15/16) was an industrial-type roller bearing, not specifically designed for marine stern tube applications. The arrangement of the drive shaft hull penetration was not in accordance with accepted marine stern tube practice. The bearing housing did not incorporate any compressible packing or sealing gland at the inboard end to prevent the entry of excessive amounts of water into the hull and was not readily available for inspection. At the start of the progressive speed trials, an initial after trim of 250mm (10inches) and an effective forward visor freeboard of 470mm (18inches) were recorded. Throughout the trials, the watertight integrity of the hull was compromised due to continuous leakage from the drive shaft bearing, shell plating fractures in way of the forward and after wheel wells and where the front wheel steering linkage penetrated the hull. The causes and locations of these fractures were most likely due to hull flexing and sudden shock loads imposed on the lightly constructed and intermittently welded bottom structure stiffening to the shell plating, when the vehicle encountered uneven and bumpy road surface conditions during on-shore driving operations, when the vehicle struck the river bottom or during recovery operations. The rates of ingress from these sources varied with the speed and trim of the vehicle and their combined accumulation was not determined during the tests, as it could not be differentiated from other concurrent and greater inflows that occurred as the trials progressed. However, the vehicle was pumped clear before the start of successive speed trials, such that the cumulative ill effects of these leaks on trim and forward freeboard were minimized for each trial condition. 1.8.3 Speed, Bow Wave and Forward Freeboard Trials Twelve trials were conducted with the hood made temporarily weathertight, with only the forward visor top gasket in the original condition it was in at the time of the occurrence. Six of these trials were conducted in calm water and the remainder, in waves created by escort vessels, simulating wave conditions likely to be met in actual passenger-carrying service (up to approximately 600mm in height). All trial speeds were determined on site by radar measurement from shore. There was no guard or stop fitted to the drive and throttle control lever of the Mercruiser I/O motor to impose a limit to the speed to which the vehicle could be driven. The top speed recorded during the trials was that obtained with the throttle lever set at the maximum available. As the speed was increased, the height of the related bow waves and the concurrent reduction in effective freeboard in way of the forward visor crossbar were observed and video-recorded from the escort vessels and from shore. The onset and subsequent increases in the flow of water over the top of the forward visor as speed was increased were also observed and recorded by video camera installed inside the forward engine compartment. Photo4. Vehicle speed of approximately 5km/h in calm water Trial speeds were progressively increased from 5km/h up to a maximum of 10.25km/h, and the related bow wave heights ranged from 90mm (8inches) to in excess of 470mm (18inches). At speeds up to about 5km/h (seePhoto4), the bow waves were comparatively small. However, as the speed was increased, the bow waves became disproportionately higher and the initial effective visor freeboard of 470mm (18inches) was significantly reduced Photo5. Vehicle speed of approximately 8km/h in calm water In calm water at a speed of 8km/h, the bow wave was some 265mm (10inches) high and the effective forward freeboard to the top of the visor was reduced to 205mm (8inches) (seePhoto5). At a speed of 9.85km/h, the bow wave was level with the visor top (seePhoto6). At the maximum speed of 10.25km/h, the bow wave partially covered the hood with water and heavy spray and overflowed the top of the forward visor. The configuration of the top of the forward visor and its attached gasket was ineffective in preventing entry of water through the space remaining between the top of the gasket and the underside of the hood. Photo6. Vehicle speed of approximately 10km/h in calm water The temporary seals over the hood sides, engine ventilation openings, and forward cowl vents were effective. However, at speeds above 10km/h, the floodwater continuously overflowed the visor top gasket and accumulated in the forward half of the vehicle. The volume of floodwater from this source was not determined due to the concurrent accumulation of water from the drive shaft bearing and hull leakage. The forward bilge pump and the two main bilgepumps were activated to clear the vehicle of all floodwater. Six speed trials conducted in water disturbed by the wakes from escort vessels showed that flooding over the top of the visor occurred sooner and at lower speeds than when in calm water conditions. At higher speeds, the amount of water and heavy spray covering the hood top occasionally washed up as far as the driver's windshield. The rate of entry and quantity of water accumulation were also greater, and the forward pump, both main bilge pumps, and the two emergency pumps were used to clear the vehicle of floodwater. Eight speed trials were conducted with the temporary weathertight seals removed from the hood sides, air ventilators and forward cowls - an operational condition similar to that at the time of the occurrence. Four of the trials were in calm water conditions and the remainder, in disturbed water. Significantly more floodwater was shipped and retained when in the unsealed condition. In disturbed water, at 8.3km/h, water overflowed the top of the visor and also downflooded through the ventilator openings, forward cowls and hood sides (seePhoto7). At 9.75km/h, the rate of downflooding increased considerably, water continuously overflowed the visor top, and a substantial quantity accumulated in the forward half of the hull (seePhoto8). The forward bilge pump, both main bilge pumps, and the two emergency bilge pumps were used to clear the vehicle of this accumulation. Photo7. Vehicle speed of 8.3km/h in disturbed water Photo8. Vehicle speed of 9.75km/h in disturbed water In disturbed water conditions, heavy spray and some solid water covered the hood, causing significant downflooding through the ventilator openings, forward cowls and hood sides. However, the greatest ingress of water was from the front, past the hood front gasket and over the top of the forward visor plate, where the overflow took on a continuous waterfall effect. At 10.24km/h, the rate of downflooding and rapid accumulation of water in the forward half of the vehicle was such that it was considered imprudent to continue at this speed for the time needed to complete the trial without the operation of all bilge pumps forward of midships. The midships bilge pumps and the forward pump were employed between all the trial runs to ensure that the hull was virtually dry and clear of floodwater at the beginning of each succeeding test. This procedure was adopted to ensure that the effective forward freeboards recorded throughout the range of speed trials were the best attainable and, as such, would be reduced in service by any unnoticed accumulation of floodwater. Any such reduction would have detrimental effects on the seakeeping characteristics of the vehicle, and either lower the speed at which the onset of downflooding through the non-weathertight hood and over the top of the visor would otherwise occur or cause higher flooding rates at the same speeds. The trial results showed that, because of bow wave effect, the forward static freeboard of 470mm (18inches) was halved at a speed of 7.6km/h. At this speed, the remaining freeboard of 235mm (9inches) would only be retained while the hull was free of all bilge or floodwater. 1.8.4 Bilge Piping Syphonic Action Trials Detailed descriptions of the pumps and the related discharge piping are included in section 1.7.1 and AppendixA. Tests were carried out to verify the onset of any water syphoning effects caused by the configuration of the discharge piping of the main bilge pump located in the starboard side drain well near midships. The discharge piping and pump did not incorporate any means to prevent the return flow of water into the hull in the event of pump malfunction while the piping was primed. The tests were carried out with the outboard end of the discharge pipe extending beyond the transom and some 305mm (12inches) below the water surface, similar to the operational condition at the time of the sinking. Tests with the discharge pipe not primed showed that syphonic action did not occur while the vehicle was stationary, moving ahead, or when going astern. Further tests carried out after the discharge piping was primed by the very brief operation of the main bilge pump showed that a steady and continuous flow of water syphoned into the hull. The inflow continued while the vehicle was stationary, moving slowly ahead and astern, and also when moving forward at 8km/h. Due to the concurrent ingress of water into the hull from other sources, the rate of flooding caused by syphonic action alone could not be determined during the water-borne tests. However, laboratory simulation of the loaded condition, with the discharge piping immersed as it was in service conditions, showed a steady flow rate of some 11.5L/min. 1.8.5 Ramp Launching Trial Trials were conducted to determine the launching characteristics of the vehicle, and to verify how much water was shipped over the top of the forward visor, through the cowl ventilators, the hood top ventilator openings and hood sides, when entering the water at various speeds. The ramp declivity at the Hull Marina was 1in6.4 and the vehicle was in a simulated fully loaded condition with an initial effective forward freeboard of 470mm (18inches) to the top of the visor plate. Launching speeds ranged from 3km/h to 11km/h and were recorded by a radar speed-measuring device. The first series of trials was carried out with all openings, except the forward end of the hood, temporarily sealed and another set was conducted with all temporary seals removed. The foam plastic gasket fitted in way of the visor top at the front of the hood was retained in its original in-service condition throughout the trials. The vehicle was initially winched into the water at minimum speed with the hood completely open. The forward freeboard at which the front of the vehicle lifted and became fully buoyant was found to be 240mm (9inches). The vehicle was retained in this position for approximately 10minutes before being hauled ashore, at which time some six litres of floodwater, which had accumulated due to hull leakage, was removed. At the lower launching speeds, the visor was not submerged and, while some spray covered the hood, there was little or no entry of water into the hull. However, at an entry speed of 11km/h, the visor was submerged, the hood and windshield were completely awash, and some 29litres of water was shipped. Photo9. Vehicle launched at 8km/h, visor submerged by approximately 25mm When the vehicle was launched at 8km/h with all temporary seals removed from the hood, the visor was just submerged, the hood briefly awash and 22litres of water was shipped (seePhoto9). When the vehicle was launched at 10km/h, the visor was submerged approximately 80mm (3inches), the hood and windshield were fully awash, and 32litres of water was shipped on board. With the vehicle in a fully loaded condition, ramp launching speeds in excess of 8km/h resulted in a significant quantity of water being shipped. In service, and in the absence of bilge high level alarms, the actual quantity of water shipped at the beginning of each tour would not be known by the operator, and its discharge was dependent on the prompt operation of the forward bilge pump. 1.8.6 Floodwater and Trim at Various Forward Freeboards A series of flooding trials were carried out to determine the progressive change of trim and the quantity of floodwater required to reduce the effective forward freeboard to the height of the bow waves generated at various speeds. The combination of the reduced forward freeboard due to accumulated floodwater and the bow wave height at a particular speed represents the flooded trim and speed at which water would overflow the top of the visor and cause continuous and major downflooding into the forward part of the hull. A forward freeboard of 470mm (18inches) and an after trim of 254mm (10inches) were recorded at the start of the flooding trials when the vehicle was free of all but a small quantity of bilge water. The vehicle was then progressively filled with water from a number of 210-litre-capacity barrels; because of the configuration of the vehicle, the floodwater was retained in the forward half of the hull. As the weight of water increased, the forward freeboard decreased and the after trim was gradually reduced. The actual quantity of water and fore and aft trim were recorded at various forward freeboards directly related to the bow wave heights determined during the preceding speed trials. While water from the seventh barrel was entering the hull, the vehicle began to trim by the bow, and a small quantity of hull leakage water accumulating at the after end of the passenger deck began to gravitate forward. Test records showed that the vehicle reached level trim when some 1375litres of floodwater was on board and that, at that time, the effective forward freeboard was 240mm (9inches). The preceding speed trials showed the bow wave height generated at 7.6km/h to be 240mm (9inches), indicating that this bow wave would be level with the top of the forward visor at this speed, when in calm water with the vehicle in a similar flooded condition. 1.9 Sinking Sequence 1.9.1 Vehicle Route and Location of Sinking Before departure, the driver was instructed that the duration of the water-borne part of the tour was to be reduced to approximately 30minutes and the LadyDuck was to briefly visit the Rideau Falls area. While returning to the HullMarina, the vehicle passed by a pleasure craft, near the Macdonald-Cartier Bridge, the operator of which remarked on the appearance and trim of the tour vehicle and took a photograph (seePhoto10). Photo10. LadyDuck approximately five minutes before sinking. Reproduced with permission. Subsequent photo-geometry and triangulation surveys carried out after the sinking verified the position of the vehicle, at the time the photograph was taken, as being approximately 250m east of the Macdonald-Cartier Bridge. These surveys and the photograph also led to the determination of the vehicle speed at that time. A review of Photo10, in conjunction with the TSB speed, bow wave, trim, forward freeboard and flooding trials, and related information, shows the following: The vehicle level trim was consistent with some 1375litres of floodwater having been shipped and retained on board, and the forward freeboard reduced to approximately 240mm (9inches). The bow wave was covering the forward end of the hood, submerging the visor plate and entering the forward cowl ventilators. The height of the bow wave at a speed of 8km/h was approximately 267mm (10inches) and exceeds the reduced forward level trim freeboard. All the passengers remained seated and not alerted at this time. An average speed of 8km/h and an elapsed sailing time of five minutes was consistent with the distance from the sinking location and the position of the vehicle when photographed. 1.9.2 Downflooding and Sinking Sequence The trials indicated how watertight integrity of the hull was not maintained. The accumulation of floodwater within the hull due to syphonic action, shaft bearing and hull fracture leaks initiated the reduction of forward freeboard before the onset of downflooding over the top of the forward visor. Furthermore, at moderate speeds, the forward freeboard was halved by a bow wave effect, leaving insufficient residual freeboard to prevent the entry of water into the hull in the event of floodwater accumulation or encounter with natural waves and wakes of other vessels in the area. The forward bilge pump, which was best situated to cope with an accumulation of water in the front of the vehicle, was inoperative and the emergency bilge pumps near midships did not stem the total inflow. As the vehicle was headed toward the HullMarina, the water accumulated in the forward end of the hull and the vehicle continued to trim more and more by the bow, causing the rate of the downflooding to accelerate. This cycle continued until forward reserve buoyancy was lost and the vehicle suddenly sank bow first. The most likely sequence of events is illustrated in Figure4. Speed=5km/h; Forward F/B=470mm; Bow wave height=89mm; Afttrim=254mm. Hull and shaft bearing leakage starts to accumulate. Port and starboard main bilge pumps activated (port-side pump inoperative). Speed = 5km/h; Forward F/B=432mm; Bow wave height=89mm; Afttrim=216mm. Hull and shaft bearing leakage continues. Starboard main bilge is fouled with solid debris. Syphonic flooding starts. Both emergency bilge pumps activated. Speed = 5km/h; Forward F/B=292mm; Bow wave height=89mm; Afttrim=38mm. Hull, shaft bearing and syphonic flooding continue. Emergency bilge pumps continue in operation. Vehicle returning to the marina. Speed=8km/h; Forward F/B=292mm; Bow wave height=267mm; Afttrim=38mm. Hull, shaft bearing and syphonic flooding continue. Bow wave starts to overflow visor and downflooding of forward end begins. Emergency bilge pumps continue in operation. MAYDAY broadcast while the vehicle is headed for shore at approximately8km/h. Bow wave=267mm. Hull, shaft bearing and syphonic flooding continue. Downflooding accumulatees as bow wave overflows visor and enters cowl vents, hood sides and air vents. Forward trim increases as vehicle settles and reserve buoyancy is reduced (seePhoto10). Downflooding becomes general with a sudden increase in forward trim as reserve buoyancy is lost and vehicle sinks rapidly by the bow. 1.10 Ministry of Transportation of Ontario Vehicle Inspection The validation tag on licence plate BP2 110 indicated an expiry date of October2001. Plate expired on 31October2001 as per permitA1699127. This vehicle was in operation in the City of Ottawa on 23June2002, at which time the registration had been expired for 7months and 23days. MTO records indicate that the registration was renewed on 31July2002 for a period extending to 31October2002. The annual inspection certificate and sticker were expired. The annual inspection sticker (No.F656674) was improperly affixed to the left lower corner of the windshield and displayed a punched date of April2001. An on-board copy of the annual inspection certificate (No.F531222), showing the vehicle description, vehicle identification number and plate number, matched the subject vehicle, except that the certificate number did not match the sticker affixed. The annual inspection sticker was found to be invalid. No evidence of a semi-annual inspection sticker was affixed to the vehicle. The rear identification lamps, side marker lamps, and clearance lamps were missing from the vehicle. Passenger seats were not securely mounted. The driver and right front passenger doors would not open. The outer body of the vehicle had been modified by the attachment of a hull sponson, effectively sealing the door in a closed position. The suspension support bracket for the third axle was found to be cracked and broken. Tires on the number three axle did not meet minimum standards for tread depth and damage to tread face. As a result of incorrect installation of the number three axle, braking efficiency was reduced. The vehicle was not equipped with at least two doors or exits. An emergency exit door was not found. The rear boarding ramp did not provide for the free egress of passengers. It was not possible to operate the road engine during the inspection of the vehicle since the engine had not been serviced after the occurrence. As a result, it was not possible to fully inspect the vehicle to minimum standards. On 18 November 2002, the licence plate was removed from the vehicle by the MTO and the vehicle was placed in unfit status as a result of its condition. 1.11 Vehicle Inspection and Certification (Marine) 1.11.1 Small Passenger Vessel Inspection Requirements As a vessel with a gross tonnage of less than5, carrying not more than 12passengers, the LadyDuckwas not subject to compliance with the Hull Construction Regulationsor the Hull Inspection Regulations, made pursuant to the Canada Shipping Act (CSA). However, it was subject to inspection pursuant to subsection316(3) of the CSA, and compliance with other regulations, including the Small Vessel Regulations (SVR), Collision Regulations (CR)and Marine Machinery Regulations. When water-borne, the vehicle was required to comply with the life-saving, safety and navigation equipment requirements of PartIV of the SVR, and the navigation lights and sound-signalling appliances necessary to comply with the CR. Such compliance was the responsibility of the owner. Additionally, as a water-borne passenger vehicle with a gross tonnage of less than 15carrying not more than 12passengers and that entered service in the spring of2001, the LadyDuck was subject to the application of the Small Vessel Monitoring and Inspection Program (SVMIP), which evolved from the Interim Small Passenger Vessel Compliance Program (ISPVCP).7 The interim guidelines in AppendixB of the SVMIP applied until the Construction Standards for Small Vessels, TP1332, was revised. The SVMIP is a voluntary compliance program in which all small vessel owners and operators are encouraged to adopt a self-monitoring inspection regime. In this manner, they can ensure that their vessels are in compliance with safety requirements that pertain to their operation, under which vessels are required: to have a first inspection and receive a Notice of Survey; to have an Annual Seaworthiness Information Report completed by the owner; to be subject to random inspection and compliance monitoring by a Marine Safety inspector, whereby a Letter of Compliance will be issued; and to be approved by TC and subject to inspection during construction. There is conflicting information with respect to the information exchanged between the owner and the regulator concerning the first inspection. However, there is no record to indicate that a formal request for inspection was made to TC. In a memorandum dated 10June1999 to all TC inspectors, the Chairman of the Board of Steamship Inspection (CBSI) indicated that this interim program was designed to promote safety and at the same time offer owners and operators a reasonable time to deal with the changes. To implement the program, a standard was developed. The memorandum stated the following: This interim standard is not highly prescriptive and allows a great deal of discretion on the part of the inspector. All decisions which meet the intention of the goals of this program will receive the support of management. The intent of the program was to permit inspectors to make decisions based on an holistic approach to safety rather than a strict application of regulations. Notwithstanding the program, an inspector is still required to apply CSA provisions that call for detention of a vessel considered to be unfit, unsafe or defective. The program was to be implemented on an as needed basis under the authority of the CBSI and the approval of the National Marine Safety Management Committee (NMSMC). This program was communicated formally to inspectors through ship safety bulletins (SSBs). Managers were briefed by regional managers with respect to the SVMIP. Managers were, in turn, expected to brief inspectors, and each office was expected to have one inspector designated as being primarily responsible for small vessels. The inspector responsible for small vessels was to be given training in the application of the SVMIP and to serve as a resource within the office for any issues pertaining to small vessels. TC established a Small Passenger Vessel Inspection Course (SPVIC) for its managers and inspectors. The pilot course was delivered in March2002 (seeAppendixF - Small Passenger Vessel Inspection Course, Manual Overview). In addition, two courses were delivered in September and December2002. As of July2003, 74managers and inspectors have attended the course and intentions are to hold additional courses. 1.11.2 LadyDuckInspection Following Previous Occurrence The LadyDuck was involved in an occurrence on 30June2001 (TSBreportM01C0033). A brief description of the occurrence and the actions taken can be found in AppendixC. Following that occurrence, examinations and tests to assess the watertight integrity of the hull were carried out independently by the TSB and TC in accordance with their respective mandates. Safety requirements, including watertight integrity, effective freeboards, trim and stability assessments, were completed to the satisfaction of TC. Related trim and intact transverse stability testing of the vehicle were conducted by TC under the aegis of the ISPVCP, with TSB as an observer. During these assessments, which were conducted with the vehicle ashore and afloat, moving at speeds somewhat less than those used in service, there was minimal ingress of water. TC, having carried out a first inspection, issued a Form S.I. 7 on 03July2001, curtailing operation of the vehicle until applicable safety requirements and standards were met. The Form S.I.7 required 18corrective measures to be addressed to meet the SVMIP first inspection requirements. The owner then contacted the TC regional office in Sarnia, Ontario, for clarification with respect to the basis for these requirements. After the discussion with the owner, the TC regional office consulted with the TC Kingston office and the manager in charge of small vessel inspections. In a letter sent to the owner on 04July2001, items on Form S.I.7 were reviewed and qualified as remains effective or no specific regulatory requirement, but this is a recommendation only by the attending inspector (seeAppendixE). The former items were all required by regulation, and the latter were safety requirements listed in the SVMIP. A copy of SSB04/2001, describing the SVMIP, was also passed to the owner at this time. Subsequently, the owner did not request to be placed on TC's SSB distribution list for future safety-related information, nor did TC add the owner to the list. TC then issued another Form S.I.7 on 20July2001, with eight additional items to be attended to by 31July2001 in order for the vehicle to be permitted to return to service. This form also advised the owner to notify TC if any abnormal seepage of water was discovered in future operation. On 01 August2001, another Form S.I.7 was issued that required two items to be added to the operator checklist for the LadyDuckas follows: the exhaust fan was to be turned on before starting the marine engine, and the forward bilge pump was to be left on automatic. On 06 May 2002, TC conducted a random inspection and compliance monitoring under the SVMIP. The inspection consisted of a visual inspection of the safety and navigation equipment, including verification of the appropriate number of lifejackets, in relation to the regulatory requirements of the SVR and the CR, and a brief trial of the vehicle on the water. The inspector did not reiterate any of the items on previous Forms S.I.7 that had been reviewed by TC and identified as recommendations, and limited the inspection to compliance with regulatory requirements. The inspection also included bilge pump testing to confirm their operation (they were turned on and heard to be running). The compliance monitoring section of Ship Inspection Certificate (SIC)99was then endorsed. At the time of this inspection, the hood of the vehicle incorporated two triangular ventilation openings. Two additional 75mm-diameter (3-inch) cowl ventilators, located at the front of the hood, above the top of the visor plate, were fitted by the owner after the inspection. This was not reported to TC, notwithstanding that SIC998 included the following: This Notice is valid for the vessel as configured on the date of inspection. No alterations to the vessel are to be undertaken without the prior approval of TC Marine Safety. 1.11.3 Small Commercial Vessel Registration/Licensing The LadyDuck was an amphibious passenger vehicle with a provincial automobile licence for highway operation and was explicitly excluded from the marine licensing requirements of PartI of the SVR (seeAppendixB for statutory and regulatory safety requirements, paragraph7(2)(d) of the SVR). In accordance with a memorandum of understanding between TC and the Department of Fisheries and Oceans (DFO),9 the departments agreed that the responsibility and authority for small vessel licensing, except for small commercial vessels, is with DFO, while responsibility and authority for ship registration, including small commercial vessels, is with TC. As an amphibious vehicle with a gross tonnage of less than15, the LadyDuckwas excluded from either licensing or registration data banks, because there was no requirement to licence an amphibious vehicle for which a provincial automobile licence for highway travel was required. 1.12 Personnel Qualifications and Certification 1.12.1 Driver Training The LadyDuck, being a vessel with a gross tonnage not exceeding 5 and carrying not more than 12 passengers, does not require a marine-certificated driver. Nevertheless, the driver held a Minister's Certificate (TC), Master Limited,10 for the operation of the LadyDiveI and the LadyDiveII, a Minister's Certificate (TC), Restricted Engineer,11 an Examiner's Certificate (TC), Master Limited/Restricted Engineer,12 for the operation of the LadyDiveIII,and a Radio Operator's Restricted Certificate, all of which were valid at the time of the occurrence. In addition, the driver held a passenger vehicle road licence from the MTO valid at the time of the occurrence. In April 2002, the driver completed a Marine Emergency Duties (MED) course at St.Lawrence College in Kingston, Ontario. A total of 45.5hours of instruction covered all sections of the approved syllabus for the MED A1 Basic Safety Course (19.5hours) and MEDA2Small Vessel Safety Course (26hours). One of the objectives of the Basic Safety Course is to ensure that seafarers are able to provide assistance in fire and abandonment emergency situations. During the course, five hours are spent on life-saving appliances and abandonment, including both theory and practical application. Course content includes training on the use of lifejackets, immersion suits, lifebuoys, liferafts and equipment, and survival craft and launching devices. Before entering the tour service, company drivers received from 12to 40hours of on-the-water instruction on board the LadyDuckand two larger amphibious passenger vehicles from company personnel. The occurrence driver had been with the company since May2001 and he estimated having received 12hours of hands-on training using the company's larger vehicles and passed a practical assessment for his Master Limited certificate, using one of those vehicles, the previous spring. Training on board the LadyDuck included familiarization with vehicle controls and operational procedures. In general, drivers practised and became familiar with the manoeuvring characteristics of the vehicle, with particular attention given to entering and leaving the water at the Hull Marina ramp. The training addressed such things as the location and operation of the vehicle's road and marine lights, horn, public address system, VHF radio, bilge pumps, fire extinguishers, stowage of lifejackets, etc. The driver of the occurrence vehicle had been given familiarization training on the LadyDuckby the company's mechanic before operating the vehicle. During training for routine preparation of the LadyDuck,drivers were verbally instructed to check that the forward and after engine fuel and other fluids were at operating levels and that all other systems were operational before beginning passenger service. Training for routine in-service operation also called for drivers to check the operation of the bilge pumps while afloat, as well as the closure of the bilge drain valves and plugs before starting service and when the vehicle was ashore and parked between tours. 1.12.2 Tour Guide Qualifications and Training This was the second summer in which the tour guide on board the vehicle had been employed by the company. The total time spent afloat was less than six months, and the tour guide had not completed MED training. The Crewing Regulationsrequire every crew member to successfully complete MED training before the completion of six months on board. Tour guides were assigned to the LadyDuck, the LadyDiveI or the LadyDiveIII. The LadyDivevessels were of different configuration, were considerably larger than the LadyDuck,with a seating capacity in excess of 40passengers, and carried a liferaft in addition to other required emergency equipment and life-saving appliances. There was no formal training program offered to guides before conducting tours on any of the amphibious vehicles and there were no written company policies or procedures on what and how to train abinitio tour guides.Tour guides were provided with a script to study, which described various land and water sites on the tour and included two references to safety: For your own safety, I just want to remind you to stay in your seat as well as keep your arms and legs in the bus at all times; and Show everyone where the lifejackets and emergency exits are situated, and that they can stand up a few at a time to take pictures. Newly hired tour guides accompanied experienced guides on tours to get a feel for how the tours should be conducted. Once the senior guide was satisfied that a tour guide had demonstrated a comfort level with delivering the script to passengers, the guide was booked to handle tours with a driver. With respect to the safety equipment on board amphibious vehicles, the tour guides were shown the location of the PFDs, lifejackets, lifebuoy, emergency exits and, in the case of the LadyDiveI and LadyDiveIII, the rear boarding ramp. In the spring of2002, the senior guide responsible for scheduling tour guides organized a meeting of those hired to date. At that meeting, the tour guides boarded the LadyDiveIand were taken on a familiarization land tour of the city. Following the tour, the guides were briefed by a company mechanic on the location of safety equipment on the LadyDiveI. The mechanic also briefed them on the location of safety equipment on the LadyDuck.This meeting was a one-off event, organized on the senior guide's initiative and not as a result of any company requirement. Tour guides were not provided with any hands-on training on how to operate any of the safety equipment and there was no reference in the script to demonstrate how to don a lifejacket. Although it was reported that demonstration of proper use of lifejackets was conducted on the larger amphibious vehicles, donning a lifejacket or a PFD was not part of the routine safety briefing on the LadyDuck. No company documentation or training was provided to crews on emergency evacuation procedures and the roles of the driver or the tour guide were not defined in the event of an abandonment of the LadyDuck. Research on emergency evacuations, especially on aircraft, has shown that strong leadership by operating personnel, involving firm direction and appropriate behaviour, serves to decrease evacuation times and reduce panic among passengers. As emergencies are not commonplace, the research also emphasizes the importance of thorough initial training and regular refresher training.13,14 1.12.3 Mechanic Qualifications and Training The company's full-time mechanic in charge of vehicle maintenance had no formal training and held no formal certification. His work-related mechanical experience was four seasons with this company. 1.13 Company's Standard Operating Procedures As a result of the occurrence involving this vehicle on 30June2001 (TSBreportM01C0033), a formal system was introduced to check and record the completion of routine vehicle preparation procedures before and during daily operation (seeAppendixC). The driver was to verify the status of the seacocks, sign the safety checklist, and have it countersigned by the tour guide or the kiosk attendant for retention ashore. The list also included the number of passengers on board, names of driver and guide, time of departure, time of entry in the water and exit from the water, and return to the kiosk. Form S.I.7 issued on 01August2001 required two items to be added to this checklist: the exhaust fan was to be turned on before starting the marine engine, and the forward bilge pump was to be left on automatic. As late as 12August2001, these items had not been added to the company checklist. A safety checklist for the day of the2002 occurrence was not recovered nor provided by the owner. However, a different document recovered from the vehicle for the day of the occurrence showed that the driver had signed off on completion of a water test (water-borne operational test) and vehicle circle check (walkaround). With the exception of the two documents above, the company had no formal standard operating procedures15 in place for the vehicles nor any written policy addressing personnel training. 1.14 Weather Conditions Between 1200 and 1700 on 23June2002, the Gatineau airport weather station recorded southwesterly winds of 7km/h, gusting to 15km/h. Thunderstorm and heavy rain showers were recorded after the occurrence from 1707 to 1913 with northwest gusts at 28km/h and with gusts to 56km/h at 1800. Since the accident occurred at about 1610, the weather was not considered to be a factor in this occurrence. 1.15 Vehicle Maintenance Records A Driver's Vehicle Inspection Report was kept by the company mechanics under the auspices of the MTO for daily inspection of the vehicle. This document recorded defects or repairs; however, there were no references to bilge pump repairs or maintenance. No other record of maintenance was available for examination after the occurrence. 1.16 Communications The vehicle was not equipped with either an automatic or manual distress radiotelephone alerting system, such as digital selective calling or an emergency position indicating radio beacon, to alert Search and Rescue (SAR) authorities. This equipment was not required by regulations for this class of vessel. Directional VHF antenna coverage of MCTS in Prescott and in Montral does not extend to the National Capital Region. Currently, there is no plan to offer such radio coverage to the Ottawa region. SAR response is available through two Canadian Coast Guard (CCG) auxiliary craft16 and any other private or commercial craft in the area that may be monitoring VHF channel16 (theemergencychannel). In 1999, the operators of the Hull (nowGatineau) 911service proposed to the CCG that a marine VHF station be incorporated in the system. However, no further action has been taken in this regard. The Ship Station (Radio) Regulations, 1999 prescribe the radio equipment to be carried by commercial vessels for distress, urgency, safety, and general communications. Ships carrying more than six passengers on a voyage within a VHF coverage area shall be equipped with one VHF radiotelephone. A proposed amendment to the regulations will require passenger ships, engaged on a voyage outside a VHF coverage area, to be equipped with radio equipment capable of establishing continuous two-way communications with a MCTS centre or a person ashore. The amendment was published in the CanadaGazette, PartI, on 12April2003, vol.137, No.15.It is anticipated that the amendment will come into force in2004. 1.17 Passenger Safety 1.17.1 Passenger Safety Briefing Safety briefing requirements are contained in PartIV of the SVR, section26.1 (for text of acts, regulations or publications, refer to AppendixB). Relevant to the LadyDuckoperation are subsections1(a), 1(d), 1(e), 1(f) and2. Before beginning the land portion of the tour, the guide briefed the passengers on procedures and safety features in French and English. This briefing advised passengers to stay in their seats and to keep arms and legs inside the vehicle at all times. Before the water-borne portion of the tour, the guide provided an additional safety briefing, in French and English, which informed the passengers that there was a lifejacket hung on a hook beside each seat, that children's lifejackets were at the back of the vehicle, and that the emergency exits were the windows along each side of the vehicle as well as the back window over the retractable stairs. During land-to-water transition, the passengers were instructed to remain seated until the vehicle had entered the water and was floating, at which time they would be able to stand up, a couple at a time, to take pictures. The lifejackets to which the tour guide referred were actually PFDs, not the approved keyhole lifejackets stowed under the passenger seats. The referred children's lifejackets were appropriate for the operation. No demonstration on how to don either the approved keyhole lifejacket or the PFD was given, nor was the passengers' attention drawn to the location of the lifebuoy or distress equipment or their methods of operation. 1.17.2 Evacuation Sequence After ordering the evacuation, and as he was moving aft to help the passengers don their PFDs, the driver was swept out of one of the window exits. To evacuate, all other occupants turned to the rear exit that had been their entrance point. The guide and one passenger, recognizing that there was a bottleneck at that exit, evacuated through the window exit beside the aftermost seat on the port side. Reportedly, the vehicle was almost vertical and the guide used the top of the awning to help pull herself out from under the water. The passenger also reported experiencing difficulty exiting through that side window. The passengers who egressed through the rear exit encountered difficulties because of the need to climb up to the opening at the top of the raised boarding ramp, the small opening provided by that exit, and minor injuries were received when passing through the opening. Common to all the accounts was the difficulty in abandonment presented by the rapidity of the sinking, the trim of the vehicle as it sank, and the bottleneck created at the rear exit. Reportedly, panic set in as passengers attempted to evacuate through the rear exit. 1.17.3 Use and Availability of Life-saving Equipment During the evacuation, one survivor held onto a PFD. Two others wore PFDs when they exited, neither of which was fastened. One of the survivors was given a lifejacket before exiting, but it was a child's lifejacket and was too small to allow the head to pass through. Another passenger, on surfacing, also used a child's lifejacket that was ineffective as a flotation device for that person. The driver and one of the passengers had not donned PFDs, the driver because he was swept overboard as he was moving to the rear of the vehicle and the other person because he had moved to the rear of the vehicle at the driver's request and was no longer within reach of the PFD at his seat. The driver was able to link his arm around a floating PFD after he was ejected from the vehicle. 1.17.4 Life-saving Equipment Requirements Life-saving equipment requirements for commercial vessels over 8m in length are contained in PartIV of the SVR, section29 (for text of acts, regulations or publications, refer to AppendixB). Life-saving equipment carried on such vessels shall provide a level of safety that is equivalent to or higher than those standards. In accordance with SSB04/2001 and the SVMIP, AppendixB, section2, paragraph (g), additional equipment is recommended, including an inflatable liferaft, inflatable platforms, and/or a suitable boat. Form S.I.7 (seeAppendixE) specified these items and, following a review by TC, they were listed as recommendations only by the attending inspector. In the SPVIC, marine safety inspectors are given instructions on life-saving equipment inspections and passenger safety briefings. Although this guidance is applicable to small passenger vessels with the emphasis on vessels of a tonnage equal to or less than 15and that carry more than 12passengers, many of the inspection items, such as lifejackets, lifebuoys and safety briefings, applied to the LadyDuck. Inspectors were required to ensure that all lifejackets were checked for strength of the outer casing, that all ties were tested for strength and completeness, that retro-reflective tape was complete and in place and that whistles were attached. Further, inspectors were to determine that lifejacket stowage was easily identified, accessible and clearly signed, and that children's lifejackets were stowed separately and readily deployable. The SPVIC instructions indicate that lifebuoys should be checked to ensure that they meet the regulations and inspectors should inform owners/operators that, although not required by law, they are encouraged to have buoyant apparatus that floats free. 1.17.4.1 Lifejackets The standard approved lifejacket to be carried on board the LadyDuck was the keyhole style, constructed to the provisions of sections 3to 7of Canadian General Standards Board (CGSB) Standard65-GP-14M.17 Twelve adult and three children's lifejackets were recovered. At the time of the occurrence, the LadyDuckhad the required number of lifejackets on board. The recovered adult lifejackets were D.O.T.-approved in1972, six years before CAN/CGSB Standard65-GP-14M came into effect. Although the lifejackets were older than the published standard, there was no requirement to replace them as long as the lifejackets were individually inspected yearly by a TC inspector and their condition pronounced adequate for the function they perform. Inspection for adequacy would include such factors as the integrity of the seams and ties, buoyancy, fabric discolouration, and required markings and whistles. A TC random inspection on 06May2002 included a verification of the number and serviceability of the lifejackets. Examination of the lifejackets after the occurrence revealed that five of the recovered adult lifejackets were not equipped with whistles and none was marked with retro-reflective tape, as is now required by CAN/CGSB Standard65-GP-14M. The markings, including the side-view sketch, were faded, in some cases to the point of being unreadable. On some of the lifejackets, the ends of the tie tapes were frayed. The three recovered children's lifejackets were D.O.T.-approved, were equipped with whistles and retro-reflective tape, and were marked appropriately. On two of the three children's lifejackets, the side-view sketch measured 55mm by 70mm, smaller than the requirement of 55mmby115mm. Photo11. Lifejacket under the seat with booster cables on top The adult lifejackets were stowed in lockers under each of the passenger seats. Examination after the occurrence revealed that it was not easy to remove the lifejackets from beneath the seats, as they tended to catch on the lip of the metal box in which they were stored. Care was necessary when removing the lifejacket during the investigation so as not to tear the shell fabric. Furthermore, since it was necessary to leave the seat to access the lifejackets, there could be considerable congestion in the single aisle. In three of the eleven lockers, there were various items (a nylon rope, a container of motor oil, and a set of booster cables) stored on top of the lifejackets (seePhoto11). The children's lifejackets were stowed at the rear of the vehicle, stacked on top of the I/O engine compartment. 1.17.4.2 Lifebuoy The LadyDuckwas not equipped with an appropriate lifebuoy. The lifebuoy on board was a 610mm-diameter (seeAppendixE), rather than the 762mm-diameter lifebuoy required by PartIV, section29 of the SVR. The 762mm lifebuoy is the largest of the lifebuoys required under the SVR and is subject to a more comprehensive list of requirements than the lesser 610mm, including the requirement for retro-reflective tape affixed at four equidistant points around the core of the lifebuoy and the attachment of a life line of not less than 9m in length. Examination of the 610mm lifebuoy revealed that, although it was stamped with TC approval number TC.143.014.045, it did not meet applicable standards as set out in both TP7325, Standards for Lifebuoys and Integral Equipment,and section8, ScheduleIII of the SVR. The lifebuoy grab lines were of varying lengths (570,590,600,and610mm) rather than the required length of not less than 610mm for each line. The lifebuoy was attached to the top of the after bulkhead above the I/O engine compartment by way of a bungy cord, the hooks of which were clasped together behind the lifebuoy. Because of the mode of attachment, it would not have been available for immediate deployment. In Form S.I.7 of 03July2001, the TC inspector requested that two 610mm-diameter lifebuoys be installed on the vehicle. A TC letter dated 04July2001 amended this request to one(1) approved lifebuoy with 9m of rope. 1.17.4.3 Personal Flotation Devices A personal flotation device is defined in the SVR as a: buoyant life-saving apparatus other than a lifejacket, that is intended to be worn by a person and that meets the standards set out in section1.3 of ScheduleIII. Section1.3 of ScheduleIII indicates that the standards for PFDs are... those set out in Canadian General Standards Board standard CAN/CGSB-65.11-M88, Personal Flotation Devices; or Underwriters Laboratories standardUL1180, Fully Inflatable Recreational Personal Flotation Devices, with the Canadian addendum. Although not required by the SVR, the LadyDuckwas equipped with 12PFDs readily available and located on hooks adjacent to the passenger seats. All PFDs recovered were size Large to XLarge, designed for chest 102cm to 122cm (40inchesto48inches) and for people over 41kilograms (90pounds). The PFDs were stamped with a label, indicating that they met standardCAN/CGSB-65.11-M88 and were approved by the CCG. At the time of the occurrence, two younger victims were wearing adult PFDs that were too large for their body size. 1.17.5 Emergency Evacuation 1.17.5.1 Emergency Exits Photo12. View from inside the vehicle of rear exit with boarding ramp in retracted position When the passengers boarded, they did so through a boarding ramp on the rear left corner of the vehicle. After all passengers boarded, the ramp was folded up by an electric winch into an opening of 1830mm high by 685mm wide (72inches high by 27inches wide) that forms a watertight seal for operation in the water. Once the ramp was secured, the opening, which was also an exit in the event of abandonment, was reduced to 788mm high (31inches) and 685mm wide (27inches) at its widest point, decreasing in width to 533mm (21inches) near the top of the exit (seePhoto12). Photo13. After boarding ramp in retracted position The ramp was articulated in that, when the ramp was retracted, its bottom two stairs folded upside down against the exterior of the vehicle. Once the vehicle was in the water, the boarding ramp could not be lowered (seePhoto13). To access this rear exit for emergency egress, passengers had to make their way along a 584mm-wide (23-inch) aisle and then make two turns to move between the rearmost seat on the port side and the front left side of the I/O engine compartment. The narrowest point of this route was 514mm (20inches). Once at the exit, passengers had to climb approximately 1016mm (40inches), with no permanent foothold to aid them, to the top of the retractable stairs to reach the available opening. There were several sharp metal corners on the top of the retractable staircase and winch assembly. The two bottom stairs of the ramp folded upside down against the exterior of the vehicle; the first of the two presented an almost slide-like property to anyone attempting to egress through this exit. Once passengers were at the top of the retractable stairs, there was no platform to step onto to facilitate egress into the water. Passengers had to find a posture to jump without any support structure to aid them in that movement, while ensuring that they cleared the hazard presented by the exterior stairs and the I/Omotor. Human engineering standards, such as Military Standard1472D (MILSTD1472D), provide guidance for design of, among other things, egress areas to ensure accommodation and compatibility by the user population. Subsection5.6.3.2 of this standard, entitled Clearance Dimensions, indicates that clearance dimensions for passageways and accesses, which must accommodate or allow passage of the body or parts of the body, shall be based upon the 95thpercentile values for applicable body dimensions. In subsection5.7.8.3, Whole Body Access, the standard states that dimensions for rectangular access side openings for body passage should not be less than a depth or height of 660mm (26inches) for light clothing and 740mm (29inches) for bulky clothing, as well as a width of 760mm (30inches) for light clothing, with 860mm (34inches) required for bulky clothing. With respect to the rear exit on the LadyDuck, while the height of the opening, 788mm (31inches), was in excess of what MILSTD1472D recommends, the width of 685mm (27inches) reducing to 533mm (21inches) fell short of being able to accommodate the full user range on which the 95th percentile values are predicated. It is instructive to note that Chapter4 of the International Code of Safety for High-Speed Craftprovides direction on accommodation and escape measurement. Regarding the dimensions of passages and accesses that form part of evacuation paths, subsection4.7.13 indicates that the dimensions should be such as to allow easy movement of persons when wearing lifejackets. There should be no protrusion in an evacuation path that could cause injury, ensnare clothing, damage lifejackets, or restrict evacuation of disabled persons. The SVMIP, AppendixB, section8, Passenger Complement, indicates that, for ClassI and ClassII vessels, the passenger complement is to be established on the basis of 610mm for each passenger on firm secured seating. Seat clearance front-to-back must be 300mm and aisles must be at least 755mm in width. ClassIII vessels, such as the LadyDuck, are only required to provide suitable seating for each person on board. A continuous fabric awning was attached to a metal frame, constructed such that its support structures formed window frames along each side of the vehicle. These frames did not obstruct egress from the openings. Each window frame was individually wrapped in fabric and secured by zippers. Eight of the window openings measured approximately 864mm by914mm (34inches by 36inches) from the interior. There were also openings, immediately behind the cab on both sides of the vehicle, that measured 495mmwide (19inches). They were irregularly shaped vertically to accommodate the downslope of the fabric awning. An additional window, located at the very rear of the starboard side, was 533mm wide (21inches) and was also irregularly shaped vertically to fit the dimensions of the vehicle. The roll-down transparent weather screens on each side of the vehicle were fitted with zippers and snap fasteners to secure them. The zippers were of two types: large zippers with double-sided slides and small zippers with single-sided slides. Of the larger openings along the side of the vehicle, four were equipped with small zippers on each side of the screen, two had large zippers on each side of the screen, and two were equipped with a combination. When the screens were zipped down, all slides on the small zippers were on the outside of the vehicle (one roll-down transparent weather screen on the starboard side near midships was zipped in the secured position on the day of the occurrence). In the closed state, none of the slides was visible or easily reachable from inside the vehicle, as they lay 50mm (2inches) below the gunwale, nor were the screen bottoms, as they were snapped to the outside of the vehicle. The sides and rear exits were not marked as emergency exits, nor was there any requirement for them to be so marked. No guidance was provided to passengers on how to use them. None of the passenger seats were equipped with a safety information card, nor was there any requirement for them to be. 1.17.5.2 Vehicle Cab Doors The cab doors on both sides of the vehicle were permanently sealed; only their windows could be raised or lowered, and provided limited means of escape from the cab in the event of abandonment. 1.18 Safety Organizations Serving Small Passenger Vessels 1.18.1 Commission des transports du Qubec In the Province of Quebec, the Commission des transports du Qubec was created to increase public safety in the areas of road, marine and rail transportation. Its mission also includes the protection of roadway heritage and management of economic activity in many transportation areas to ensure service availability and quality. Any person who offers passenger transportation services on water for commercial purposes must hold a permit issued by the Commission. That permit provides assurance that the holder has the knowledge and expertise required to offer such services, that the holder has obtained the minimum insurance coverage indicated in the regulations, and that the vessel meets safety requirements. The carrier must provide a crew with applicable knowledge and experience to allow them to make competent use of the permit they have received. Carriers must also provide the Commission with an insurance certificate, stating that every one of the ships for which they requested a permit is covered by a maritime civil liability insurance policy with a guaranteed minimum limit of five million dollars. This applies to every vessel with a gross tonnage over 5or a capacity of over 12passengers. Carriers must also meet the requirements of TC by providing a ship inspection certificate issued by TC. For vessels with a smaller capacity, a letter of compliance, issued by a maritime expert recognized by TC or one of its inspectors, is required along with a note confirming the capacity of the ship and the competence of the crew to undertake commercial activities over a given territory. The maritime civil liability insurance certificate for ships with a gross tonnage equal to or below5 or a capacity of not more than 12passengers must have a minimal guarantee of one million dollars per vessel. Maintaining crew qualifications and ship quality and holding an insurance policy are the first conditions required for a maritime passenger transport permit covering all vessel categories. Since the Rglement sur le transport maritime de passagers came into force in1998,the Commission has registered 252vessels, including 86 vessels of a gross tonnage equal to or below5, and 106vessels with a capacity of not more than 12passengers. Because the vehicle was registered in Ontario, the LadyDuckwas not required to hold a permit issued by the Commission. No parallel organization exists in Ontario. 1.18.2 Operator Vessel Safety Standards for Whale Watching Vessels In the Province of British Columbia, in collaboration with TC and the Whale Watch Operators Association Northwest, safety standards have been put in place for the operation of whale watching vessels of a gross tonnage equal to or below5 and carrying not more than 12passengers that operate from the Greater Victoria city area. The standards are composed of the following three sections: Standards for Victoria area whale watching companies operating vessels of gross tonnage less than5 and carrying12or fewer passengers These standards were submitted to TC with the intention of developing them into a TC standard. They contain specific safety requirements that address vessel construction and seaworthiness, safety equipment, communication equipment, navigation equipment, and operator proficiency requirements. Code of Conduct The Code was established by the Whale Watch Operators Association Northwest as a guideline for operating procedures to ensure the safety and comfort of the occupants aboard rigid hull inflatable whale watch vessels. The Code is not intended to be part of the TC standard. It contains, interalia,guidelines that address occupant safety, weather concerns, areas of adverse conditions, speed limits, and operator etiquette. Training syllabus for the whale watching operators proficiency Victoria Harbour and approaches for commercial vessels of gross tonnage not exceeding5 and that do not carry more than 12passengers The syllabus was prepared by TC (Victoria) in consultation with the Whale Watch Operators Association Northwest and is not intended to be part of the TC standard. It provides an outline of the information considered essential to the safe execution of an operator's duties. The material is also intended to be a guide for the operator's examination. No similar standards applicable to the LadyDuckexist in the Province of Ontario. 1.18.3 Canadian Passenger Vessel Association The Canadian Passenger Vessel Association (CPVA) is composed of more than 120commercial passenger vessel owners and operators who offer transportation services to the public in Canadian waters. The purpose of the CPVA is to exchange ideas and information among the membership and to maintain a dialogue with the regulatory agencies on rules that govern the industry. The CPVA promotes the ideals of marine safety among its membership and a continuing quality of services for the travelling public. The CPVA has developed a safety training course for non-certificated personnel serving on board seasonal passenger vessels. The objective of this course is to provide a ship-specific safety course that may be given by a certificated officer or other approved person at a one-day session at the outset of the season. This course is a substitute for MED A1 for crew that had received provisional approval on 28March2002, and had received final approval by TC two days before this occurrence. It includes theoretical and practical elements on emergencies, emergency response, life-saving appliances, abandonment, survival, fire fighting, accidents on board, and rescue. The operator of the LadyDuck was not a member of the CPVA.